Google DeepMind Debuts Project Genie, AI Tool for Creating Interactive Worlds

Google DeepMind has launched Project Genie, an experimental research prototype that allows users to create and explore interactive digital worlds using artificial intelligence, the company announced last week.

The tool is initially rolling out to Google AI Ultra subscribers in the United States and represents a step toward advancing AI “world models” that simulate dynamic environments in real time.

Project Genie is powered by DeepMind’s Genie 3 model, along with Google’s Nano Banana Pro and Gemini AI technologies.

"In August, we previewed Genie 3, a general-purpose world model capable of generating diverse, interactive environments. Even in this early form, trusted testers were able to create an impressive range of fascinating worlds and experiences, and uncovered entirely new ways to use it. The next step is to broaden access through a dedicated, interactive prototype focused on immersive world creation," Google said in a blog post.

Users can generate living, navigable worlds by providing text prompts or uploading images, sketching scenes before entering them and then exploring environments that evolve as they move.

The prototype supports three main functions: world sketching, world exploration, and world remixing, letting creators build on existing scenes or remix them into new experiences.

In these AI-generated environments, users can traverse terrain, adjust characters and perspectives, and even download videos of their explorations. As Project Genie generates the path ahead in real time, it simulates elements like movement and basic physics, although DeepMind notes that generated worlds may not fully adhere to real-world physics or image prompts in this early form.

Google said the project is experimental and part of its broader research into AGI-scale world models capable of reasoning about environments, actions and consequences.

Nonetheless, the company emphasised that the technology is still in development, with known limitations such as generation quality and interaction controls. Access will expand to more territories over time as Google gathers feedback on how people use and create within these AI-driven worlds.